Other use cases of the repository include: exposing technology based on deep learning, using pre-trained nets as feature extractors, and building nets using off-the-self architectures and pre-trained components. The goal of the repository will be to curate and publish models in a way that is easily accessible and consumable in addition to providing its own trained models. The team decided to launch the Wolfram Neural Net Repository because “training state-of-the art neural nets often requires huge datasets and significant computational resources that are inaccessible to most users.” With a publicly available repository, users can access the most current architectures and pre-trained nets with “thousands of hours of computation time on powerful GPUS.”
And deep learning will no doubt play an important role in our continuing mission to make human knowledge computable,” the Wolfram team wrote in a post. This has made possible a whole new set of Wolfram Language functions, such as FindTextualAnswer, ImageIdentify, ImageRestyle and FacialFeatures. Fortunately, the Wolfram Language now has a state-of-the-art neural net framework (and a growing tutorial collection). “Neural nets have generated a lot of interest recently, and rightly so: they form the basis for state-of-the-art solutions to a dizzying array of problems, from speech recognition to machine translation, from autonomous driving to playing Go. The Wolfram Language neural network framework includes models for automated machine learning, representation, operations, basic layers, recurrent layers, sequence-handling-layers, training optimization layers, and managing data and training. The Wolfram Neural Net Repository builds on the company’s Wolfram Language neural framework to store neural net models and enable the immediate use for evaluation, training, visualization and transfer learning.
#WOLFRAM PLAYER PLUGIN SOFTWARE#
The software company Wolfram Research is launching a public repository for trained and untrained neural network models.
0 kommentar(er)
